US8660329B2 - Method for reconstruction of a three-dimensional model of a body structure - Google Patents

Method for reconstruction of a three-dimensional model of a body structure Download PDF

Info

Publication number
US8660329B2
US8660329B2 US12/601,975 US60197507A US8660329B2 US 8660329 B2 US8660329 B2 US 8660329B2 US 60197507 A US60197507 A US 60197507A US 8660329 B2 US8660329 B2 US 8660329B2
Authority
US
United States
Prior art keywords
subject
knowledge base
reconstruction
preliminary solution
specific
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active, expires
Application number
US12/601,975
Other languages
English (en)
Other versions
US20100174673A1 (en
Inventor
Wafa Skalli
Ludovic Humbert
David Mitton
Jean Dubousset
Jacques de Guise
Benoit Godbout
Stefan Parent
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Ecole de Technologie Superieure
Centre National de la Recherche Scientifique CNRS
Ecole National Superieure dArts et Metiers ENSAM
Original Assignee
Ecole de Technologie Superieure
Centre National de la Recherche Scientifique CNRS
Ecole National Superieure dArts et Metiers ENSAM
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Ecole de Technologie Superieure, Centre National de la Recherche Scientifique CNRS, Ecole National Superieure dArts et Metiers ENSAM filed Critical Ecole de Technologie Superieure
Publication of US20100174673A1 publication Critical patent/US20100174673A1/en
Assigned to ECOLE DE TECHNOLOGIE SUPERIEURE reassignment ECOLE DE TECHNOLOGIE SUPERIEURE ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS). Assignors: DE GUISE, JACQUES, GODBOUT, BENOIT, PARENT, STEFAN
Assigned to ECOLE NATIONALE SUPERIEURE D'ARTS ET METIERS (ENSAM), CENTRE NATIONAL DE LA RECHERCHE SCIENTIFIQUE (CNRS) reassignment ECOLE NATIONALE SUPERIEURE D'ARTS ET METIERS (ENSAM) ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS). Assignors: HUMBERT, LUDOVIC
Assigned to ECOLE NATIONALE SUPERIEURE D'ARTS ET METIERS (ENSAM) reassignment ECOLE NATIONALE SUPERIEURE D'ARTS ET METIERS (ENSAM) ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS). Assignors: SKALLI, WAFA, DUBOUSSET, JEAN, MITTON, DAVID
Application granted granted Critical
Publication of US8660329B2 publication Critical patent/US8660329B2/en
Active legal-status Critical Current
Adjusted expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/50—Depth or shape recovery
    • G06T7/55—Depth or shape recovery from multiple images
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T17/00—Three dimensional [3D] modelling, e.g. data description of 3D objects
    • G06T17/10—Constructive solid geometry [CSG] using solid primitives, e.g. cylinders, cubes
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2200/00—Indexing scheme for image data processing or generation, in general
    • G06T2200/08—Indexing scheme for image data processing or generation, in general involving all processing steps from image acquisition to 3D model generation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30004—Biomedical image processing
    • G06T2207/30008—Bone
    • G06T2207/30012—Spine; Backbone

Definitions

  • the instant invention relates to a method for reconstruction of a three-dimensional model of a body structure.
  • the instant invention is related to a method for reconstruction of a three-dimensional model of a body structure, said structure comprising a plurality of objects, wherein
  • a knowledge base of said structure comprising a knowledge base of each object of said structure
  • a preliminary solution of at least a first object of the structure is obtained from said knowledge base of the object.
  • FR 2 856 170 describes such a method for reconstruction, which enables to obtain a three-dimensional subject-specific reconstruction of an osteo-articular, such as the human spine, from few calibrated radiographs. Although this process is quick, reliable and limits the intervention of the human operator, there is a continued need to obtain more automatic reconstruction methods. There is a need to limit human intervention, so that reproducibility of the reconstruction method is enhanced, and that its access to untrained personal, not having high knowledge in anatomy, can be facilitated.
  • the instant invention has notably for object to mitigate at least one of these drawbacks.
  • the invention relates to a method for reconstruction of a three-dimensional model of a body structure of a patient, said structure comprising a plurality of objects, wherein
  • a knowledge base of said structure comprising a priori knowledge of the structure, previously established from structures of the same type as the structure to be reconstructed, said knowledge base of structure comprising a knowledge base of each object of said structure, the knowledge base of each object thus comprising a priori knowledge base of the respective object, previously established from objects of the same type as the respective object, the knowledge base of said structure being adapted to estimate, from a parameter of any object of the structure, another parameter for at least an object of the structure,
  • the preliminary solution of at least a first portion of an object is selected from the preliminary solution of the structure
  • At step (f) at least a three-dimensional subject-specific reconstruction of at least a second portion of object is obtained based on the modified preliminary solution of the first portion of object obtained at step (e) and on the knowledge base of structure.
  • a subject-specific reconstruction of at least a portion of an object is obtained based on the modification of the preliminary solution of a portion of a first object.
  • This enables to quickly obtain an accurate reconstruction of at least a sub-part of the structure with a minimal human user intervention.
  • accurate subject-specific reconstructions of objects for which few detection data are available could be obtained based on fitting the subject-specific reconstruction for other objects on detection data of better quality.
  • the method will be particularly powerful when the structure comprises many objects, such as the spine or the whole skeleton.
  • Such a method can be relied upon by medical end users such as surgeons for diagnostic and/or preoperative planning for example. It could be quick enough to enable for example real time reconstruction for computer-assisted navigation applications based on intra-operative detection data during surgery.
  • the second set of parameters for the subject-specific reconstruction of the second portion of object is calculated by statistical inference on the knowledge base of structure using the first set of parameters of the first portion of object as an estimator;
  • the invention relates to a computer program product comprising instructions for causing a programmable unit to perform the above methods when executed on said programmable unit.
  • the invention relates to a computer program product comprising a knowledge base of a body structure, said structure comprising a plurality of objects, said knowledge base being for use in a method for reconstruction of a three-dimensional model of a subject-specific body structure based on subject-specific data, said knowledge base comprising a priori-knowledge of the body structure, previously established from structures of the same type as the subject-specific structure, said knowledge base of structure comprising a knowledge base of each object of the structure, said knowledge base of structure being adapted to estimate, from a parameter of any object of the structure, another parameter for at least one object of the structure, the knowledge base of each object being adapted to estimate, from a parameter of the respective object, a three-dimensional subject-specific reconstruction of at least a portion of the respective object.
  • FIG. 1 is a perspective view of an example of a reconstruction apparatus
  • FIG. 2 is a schematic view of an example of a database of a structure to be reconstructed
  • FIG. 3 is a perspective view of an example of a generic representation of an object of a structure
  • FIG. 4 is a schematic view of a computer screen showing detection data of the structure
  • FIG. 5 is a schematic perspective view of a preliminary solution for a subject-specific reconstruction of the structure for which detection data was shown on FIG. 4 ,
  • FIG. 6 a is a schematic partial view of a computer screen showing both the detection data for each object (in dotted lines), as well as the projection of the preliminary solution for the subject-specific reconstruction of FIG. 5 in the plane of the detection data,
  • FIG. 6 b is a view corresponding to FIG. 6 a of the fitting of the preliminary solution for the subject-specific reconstruction for vertebra L1 to detection data for L1, by rigid displacement,
  • FIG. 6 c is a view corresponding to FIG. 6 a after the preliminary solution for the subject-specific reconstruction for vertebra L1 has been fitted to detection data for L1,
  • FIG. 7 a is a view corresponding to FIG. 6 a for an other body structure
  • FIG. 7 b is a view corresponding to FIG. 7 a of the fitting of the preliminary solution for the subject-specific reconstruction for vertebra L1 to detection data for L1 by non-rigid deformation
  • FIG. 7 c is a view corresponding to FIG. 7 a after the preliminary solution for the subject-specific reconstruction for vertebra L1 has been fitted to detection data for L1,
  • FIGS. 8 a - 8 c are views corresponding to FIGS. 6 a - 6 c for a third embodiment of the invention.
  • FIG. 9 is a schematic perspective view of another possible embodiment for an installation.
  • FIG. 1 shows a radiographic apparatus 1 for three-dimensional reconstruction, the apparatus comprising a moving frame 2 displaceable under motor drive along vertical guides 3 in both directions of translation 3 a.
  • the frame surrounds a field of observation 4 in which a patient P can be placed, e.g. standing, for observing an osteo-articular structure of the patient when in the standing position, which can be important for patients suffering from scoliosis, for example.
  • the moving frame 2 carries a first radiological source 5 and a first detector 6 which is placed facing the source 5 beyond the field 4 , and which comprises at least one horizontal line 6 a of detector cells.
  • the detector 6 may be a gas detector responsive to low doses of radiation, e.g. as described in documents FR-A-2 749 402 or FR-A-2 754 068. Naturally, other types of detectors may optionally be used in the context of the present invention.
  • the radiological source 5 is adapted to emit ionizing radiation, in particular X-rays, suitable for being detected by the detector 6 in an image-taking direction 7 that is antero-posterior relative to the patient P, the rays passing through a horizontal slit 8 made through an aiming mask 9 such as a metal plate in order to generate a horizontal beam 10 of ionizing radiation in the field of observation 4 .
  • ionizing radiation in particular X-rays
  • the moving frame 2 also carries a second radiological source 11 similar to the source 5 and a second detector 12 similar to the detector 6 , disposed facing the source 11 beyond the field 4 , and comprising at least one horizontal line 12 a of detector cells.
  • the radiological source 11 is adapted to emit ionizing radiation in a image-taking direction 13 that is lateral relative to the patient P, passing through a horizontal slit 14 formed in an aiming mask 15 such as a metal plate in order to generate a horizontal beam 16 of ionizing radiation in the field of observation 4 .
  • radiological sources and detectors there could be more than two radiological sources and detectors, and the image-taking directions of these various radiological sources could, where appropriate, be other than mutually perpendicular, and they need not even be horizontal.
  • the two detectors 6 , 12 are connected to a computerized system 37 or some other electronic control system fitted with:
  • the microcomputer 37 may also be connected to the motor-driven drive means (not shown) contained in the guide 3 , and to the sources 5 and 11 , so as to control vertical displacement of the frame and the emission of ionizing radiation.
  • the reconstruction method which is described after is a reconstruction method of the human spine using a knowledge base of spines.
  • this method could be used for reconstruction of any structure of the body such as osteo-articular structures comprising a plurality of bony objects, such as for example, the upper limb, the lower limb, the hip, or even part or totality of the skeleton, when a knowledge base of the structure to be reconstructed is provided.
  • the structure could thus comprise as few as two objects such as for example a femur and a tibia.
  • the knowledge base 21 of the structure comprises a database of each of the objects of the structure to be reconstructed.
  • the knowledge base 21 comprises a database of each of the vertebral bodies.
  • the database of the structure 21 comprises the database 22 c1 of the first cervical vertebra, a database 22 C2 of the second cervical vertebra, . . . a database of each other vertebra, (e.g. a database 22 Ti of the i th thoracic vertebra for the i th vertebra), . . . , and a database 22 L5 of the fifth lumbar vertebra.
  • the database for the object can be constructed as or from a data obtained from objects similar to the object to be reconstructed.
  • the database may contain data relating to particular reference marks on objects of the same type of the object to be reconstructed, acquired beforehand, for example by computer tomography.
  • the database of the i th vertebra Ti contains the positions of characteristic points P 1 , . . . P 23 for the i th vertebra of each of a plurality of patients, characteristic lengths D 1 . . . D 8 for each vertebra, as shown in FIG. 3 , segments, straight lines or arcs that are characteristics of the object, and/or outlines and edges of these particular vertebra.
  • the coordinates of characteristic points or lines may be expressed, for example, in a local X, Y, Z frame of reference.
  • the axis Z corresponds to the axial direction of the vertebral column
  • the axis X is determined so as to define the antero-posterior plane of the vertebra 20 along with axis Z
  • the axis Y being perpendicular to the above-mentioned axes X and Z.
  • the origin O of the frame of reference is disposed in the middle between the two axial end faces of the main “tubular” portion of the vertebra, the origin O also being positioned so that the axis Z passes through the upper axial face of the main portion of the vertebra at a reference feature P 1 such that the distance between said reference feature P 1 to the front end P 7 of said axial face is equal to about two-thirds of the total distance between the front and rear ends P 7 and P 8 of the antero-posterior section of said top axial face.
  • the database of vertebra Ti thus comprises means R Ti for calculating, from one or more estimators for the vertebra being currently reconstructed, the coordinates of points P 1 -P 23 and/or other geometric data, by statistical inference on the database of vertebras T i . It is also possible to establish a subset of the database of vertebrae belonging to healthy individuals or to individuals suffering from scoliosis, and similarly it is possible to characterize each vertebra as a function of the weight, the size, the age, or any other type of parameter concerning the individual that is deemed to be necessary.
  • the knowledge base may comprise a mathematical model constructed from the previously acquired data.
  • the knowledge base may include statistical data R Ti (means, variances, . . . ) for each parameter of the knowledge base, or indeed mathematical equations for determining from the knowledge base of a given vertebra T i , the positions of the characteristic points for a subject-specific reconstruction on the basis of values of estimator parameters for said object.
  • the coordinates of the control points will be parameterized by functions of these parameters.
  • the knowledge base can also contain data relating to the position of a given vertebra in the vertebral column of the subject submitted to the CT-scan to enter its vertebra into the database, such as, for example, the angular orientation of the vertebra and the curvature of the spinal column at the level of that vertebra.
  • the knowledge base will comprise geometrical information which is known a priori, such as a prerequisite that a given vertebral endplate will be orthogonal to the spinal center line.
  • a generic model of each vertebra is also available, for example established from the database as an average vertebra for that vertebral level.
  • the generic model is defined as a mesh of several hundred to several hundred thousand points of a mean vertebra T i .
  • a generic model can be provided for each type of vertebra, for example.
  • the knowledge base of the structure further comprises relationships between the above-described models of objects.
  • relationships between vertebra T i and vertebra will be identified on FIG. 2 by reference R Ti-Ti-1 . It is not necessarily only neighbouring vertebras which have pre-established relationships.
  • the relationship R C2-Ti describes a relationship between vertebra C 2 and vertebra T i .
  • a knowledge base of vertebras C2 and a knowledge base of vertebras Ti are established from detection data previously obtained from whole spines. Relationships linking characteristic dimensions or coordinates of these two vertebras are also linked together for each spine.
  • the knowledge base of the structure will comprise means for calculating specific coordinates or geometrical characteristics of a given vertebra Ti by statistical inference on the knowledge base of structure based on an estimator which is provided by estimating a geometrical characteristic of another vertebra T k .
  • coordinates of control points P 1 -P 23 of each given vertebra could be provided as parameterized by functions of geometrical parameters for both the given vertebra and neighbouring vertebra.
  • the knowledge base of each object comprises a priori knowledge of this object
  • the knowledge base of the structure comprises a priori knowledge of the structure in the form of a priori knowledge of each object and of relationships between the objects.
  • the microcomputer 37 is used initially to take two radiographic images of the patient P by causing the field of observation 4 to be scanned by the beams 10 and 16 of ionizing radiation over a height corresponding to the structure of the patient that is to be observed, for example the spine and the pelvis, or indeed the entire skeleton.
  • the frame is preferably displaceable over a height of not less than 70 centimeters (cm), and preferably over at least one meter.
  • two calibrated digital radio-graphic images of the portion of the patient under examination are stored in the memory of the microcomputer 37 , for example an antero-posterior image and a lateral image respectively, which images can be viewed on the screen 19 of the microcomputer, as shown in FIG. 4 .
  • Bidimensional detection data of the structure such as an antero-posterior and a lateral radiographs obtained by the apparatus of FIG. 1 , is displayed on a screen 19 of the computerized system 37 as shown on FIG. 4 .
  • detection data of the objects are clearly shown on FIG. 4 , which is a schematic representation, it should be understood that detection data such as radiographs, do not necessarily provide with such a clear representation of the structure. This is illustrated by the presence of the arm and lung on the lateral view which hides most of the upper spine in this view.
  • Projections of the objects of the structure of the patient are identified by reference number 23 b on the frontal image (right side of FIG. 4 ), and 23 a on the lateral image (left side of FIG. 4 ).
  • the projections 23 a , 23 b of all of the objects of the structure form all together the projection of the structure.
  • the preliminary solution could for example be obtained as follows:
  • the length of the three-dimensional central line of the spine is calculated, and the positions of the subject-specific reconstruction of the intermediate objects are calculated as being regularly spaced along said line, based on the position of the preliminary solution of the subject-specific reconstruction of the first (lowermost or uppermost) object as a parameter: this (these) positions will be used to determine the positions for the preliminary solution of the subject-specific reconstruction of the intermediate objects.
  • the orientation of the preliminary solution of the subject-specific reconstruction of each object is calculated from the orientation of the tangent to the three-dimensional central line at the calculated position, and of the relative orientations of tangent to the three-dimensional central line above and under the calculated position.
  • the positions of the preliminary solution of the subject-specific reconstruction of each object could be determined in other ways.
  • the subject-specific reconstructions of objects will not be regularly spaced along the three-dimensional central line, but will be placed irregularly along this line.
  • These positions could for example be estimated from a non-linear function.
  • This non-linear function will for example be calculated beforehand by taking into account a priori knowledge of the structure.
  • the non-linear function will take into account the a priori spacing of consecutive vertebra.
  • the relative positions could be obtained by statistical inference on the knowledge base 21 of structure, taking as estimators the length of the three-dimensional line, and the three-dimensional orientation of the uppermost and lowermost vertebral plates, and/or other parameters of the patient such as height, weight, etc. In order to do so, a priori knowledge would be provided by the relationships R c1-c2, . . . , R L4-L5 of the knowledge base 21 of structure.
  • orientations of each of the preliminary solutions of the subject-specific reconstructions of the objects could also be obtained likewise by statistical inference on the knowledge base of structure using the orientation of a previously reconstructed object as an estimator.
  • the preliminary solution of the subject-specific reconstruction itself for each object 28 is obtained by calculating the three-dimensional coordinates of points of the subject-specific reconstruction of each object based on the above calculated positions and orientations, such as for example, by statistical inference on the knowledge base of each respective object, using these positions and orientations as estimators.
  • the positions in three dimensional space of points P 1 -P 23 will be calculated from the knowledge base of the vertebra T i , using the estimated position and orientation as an estimator by statistical inference on the database of T i .
  • the generic model of T i is deformed or displaced to fit with the coordinates of points P 1 -P 23 in 3D space.
  • the obtained three-dimensional preliminary solution for a subject-specific reconstruction of each object 28 is represented on FIG. 5 . They all together constitute the preliminary solution of the subject-specific reconstruction 27 of the structure.
  • FIG. 5 obtained as described above, could be considered as a preliminary solution for implementing the steps which will be described below in relation to FIGS. 6 a - 6 c , 7 a - 7 c , or 8 a - 8 c.
  • preliminary solution it is meant a subject-specific reconstruction having a given accuracy, and that a method will be applied to this “preliminary solution” to obtain a more accurate subject-specific reconstruction.
  • This more accurate subject-specific reconstruction could still itself be considered as a “preliminary solution” for a further implementation of the method in view of obtaining a further refined construction.
  • FIG. 6 a schematically represents a part of the left window of the computer display 19 , wherein the detection data of the structure is partly represented in dotted lines. It should be observed that this detection data could for example be provided in the form of a radiograph, and that the outline of the projected object is not necessarily as clear as shown on FIG. 6 a in the application of the herein described method.
  • a projection of the preliminary solution for the subject-specific reconstruction of the structure onto the detection plane is also displayed on FIG. 6 a in continuous lines. This projection is for example obtained by simulating the tangency of X-rays on the subject-specific reconstruction of the structure. As explained above, this preliminary solution is for example the reconstruction of FIG. 5 . As shown on FIG.
  • the preliminary solution is not necessarily very well fitted to the detection data.
  • one of the objects is selected.
  • vertebra L1 is selected because it is estimated or calculated that its initial solution is very distant from the corresponding detection data, or for any other reason, such as that very accurate detection data are provided for L1.
  • the subject-specific reconstruction of the first object (L1) of the preliminary solution is modified. For example, a first point A 1 is identified on the projection of the preliminary solution for L1, and a similar point in the detection data is identified as A 2 on the display screen.
  • This identification on the lateral view could be performed either by a user with input means 18 , or by computerized numerical treatment of images. Possibly, the same steps are accomplished, for the same vertebra, on the frontal view displayed in the right window of the computer screen (not shown).
  • the subject-specific reconstruction for L1 is obtained by modifying the preliminary solution so that points A 1 and A 2 are brought into concordance.
  • the modification to the preliminary solution could for example be, in this example, a rigid displacement of the subject-specific reconstruction of L1 by applying a translation and/or a rotation in 3D space to the preliminary solution.
  • the displacement in 3D space of the initial solution for L1 which enables bringing in correspondence A 1 with A 2 , and simultaneously similar points in the frontal view, is calculated.
  • the subject-specific reconstruction of at least another object is automatically obtained from the above-described modification.
  • the subject-specific reconstruction for T12 is obtained, from the preliminary solution for T12 shown on FIG. 6 a , by modifying this preliminary solution on the basis of the modifications applied to the subject-specific reconstruction of L1.
  • the knowledge base of structure 21 is used, and in particular the relationship R L1-T12 between vertebrae L1 and T12. For example, the new position and/or orientation of the subject-specific reconstruction of L1 in FIG.
  • 6 b will provide a parameter of object L1, to be used as an estimator for obtaining another estimator of object T12 such as the position and/or orientation and/or shape of the subject-specific reconstruction of T12 by statistical inference on the knowledge base of structure 21 . Possibly, this calculation will also take into account the position and/or orientation of the preliminary solution for vertebra T11, using relationship R T11-T12 between vertebra T 11 and T 12 in the knowledge base of structure. Possibly, other objects will also be taken into account.
  • the subject-specific reconstruction in three-dimensional space of T12 is obtained from the knowledge base of T12, for example by statistical inference on this knowledge base using the determined characteristic (position, orientation, parameter of shape) as an estimator.
  • This step which is described for vertebra T12 based on the fit of vertebra L1 could simultaneously be performed for all the other objects to be reconstructed in order to obtain the subject-specific reconstruction of the structure, a projection of which in the detection data plane is shown on FIG. 6 c .
  • this method will provide more accurate subject-specific reconstruction for objects neighbouring the subject-specific reconstruction of the originally modified object than for more remote objects.
  • this method could be performed iteratively, on different objects, using a previously obtained reconstruction as a “preliminary solution”, until a satisfying subject-specific reconstruction for the structure as a whole is obtained.
  • the knowledge base of the structure has means to estimate a parameter for at least a second object for each of the first objects.
  • FIG. 7 a schematically represents a part of the left window of the computer display 19 , wherein the detection data of another structure is partially represented in dotted lines. Further, a projection of the preliminary solution for the subject-specific reconstruction of the structure onto the detection plane is also displayed on FIG. 7 a in continuous lines.
  • one of the objects is selected. For example, vertebra L1 is selected.
  • the subject-specific reconstruction of the first object (L1) of the preliminary solution is modified. For example, two points A 3 , A 4 are identified on the projection of the preliminary solution for L1, and similar points in the detection data are identified (not shown) on the display screen. Possibly, the same steps are accomplished, for the same vertebra, on the frontal view displayed in the right window of the computer screen.
  • the subject-specific reconstruction for L1 is obtained by modifying the preliminary solution so that points A 3 and A 4 are brought into concordance with the corresponding points in the detection data.
  • the modification to the preliminary solution could for example be, in this example, linear deformation of the subject-specific reconstruction of L1 by applying a homothetic transformation and/or a rotation to the initial solution.
  • the subject-specific reconstruction of at least another object is automatically obtained from the above-described modification.
  • the subject-specific reconstruction for T12 is obtained, from the preliminary solution for T12 shown on FIG. 7 a , by modifying this initial solution on the basis of the modifications applied to the subject-specific reconstruction of L1.
  • the knowledge base of structure 21 is used, and in particular the relationship R L1-T12 between vertebrae L1 and T12.
  • the new position, orientation and shape of the subject-specific reconstruction of L1 will provide a parameter of object L1 to be used as an estimator for the position, orientation and shape of the subject-specific reconstruction of T12 by statistical inference on the knowledge base of structure 21 .
  • the subject-specific reconstruction in three-dimensional space of T12 is obtained from the knowledge base 22 T12 of T12, for example by statistical inference on this knowledge base using the determined characteristics (position, orientation and shape parameter) as estimators.
  • This step which is described for vertebra T12 could be automatically reproduced for all the objects to be reconstructed in order to obtain the subject-specific reconstruction of the structure, a projection of which in the detection data plane is represented on FIG. 7 c.
  • this method will provide an accurate subject-specific reconstruction for objects neighbouring the subject-specific reconstruction of the originally modified object. This method could be performed a given number of times, on different objects, until a satisfying subject-specific reconstruction for the structure as a whole is obtained.
  • the knowledge base is adapted to estimate a parameter of a portion of an object, for example the lower endplate of vertebra T12, from a parameter of a portion of an object, for example the upper endplate of vertebra L1, for example based on a relationship R L1u-T12l between these two parameters.
  • FIG. 8 a schematically represents a part of the left window of the computer display 19 , wherein the detection data of another structure is partially represented in dotted lines. Further, a projection of the preliminary solution for the subject-specific reconstruction of the structure onto the detection plane is also displayed on FIG. 8 a in continuous lines.
  • a portion of one of the objects is selected. For example, the upper endplate of vertebra L1 is selected.
  • the subject-specific reconstruction of the first portion of object (upper endplate of L1) of the preliminary solution is modified. For example, one point A 5 is identified on the projection of the preliminary solution for the upper endplate L1, and a similar point in the detection data is identified as A 6 on the display screen. Possibly, the same steps are accomplished, for the same vertebra, on the frontal view displayed in the right window of the computer screen.
  • the subject-specific reconstruction for the upper end plate of L1 is obtained by modifying the preliminary solution so that point A 5 is brought into concordance with the corresponding point A 6 in the detection data.
  • the modification to the preliminary solution could for example be, in this example, by manually displacing the initial solution of the subject-specific reconstruction of the upper endplate of L1.
  • the subject-specific reconstruction of at least another portion object is automatically obtained from the above-described modification. For example, if one considers the lower vertebral endplate of vertebra T12 as the second portion of object, the subject-specific reconstruction for this second portion of object is obtained, from the preliminary solution for this second portion of object shown on FIG. 8 a , by modifying this initial solution on the basis of the modifications applied to the subject-specific reconstruction of the upper endplate of L1. In order to do so, the knowledge base of structure 21 is used and in particular the relationship R L1u-T12l between the upper endplate of vertebrae L1 and the lower endplate of T12.
  • the new position, orientation and shape of the subject-specific reconstruction of the upper endplate of L1 will provide a parameter of this portion of object to be used as an estimator for the position, orientation and shape of the subject-specific reconstruction of the lower vertebral endplate of T12 by statistical inference on the knowledge base of structure 21 .
  • the subject-specific reconstruction in three-dimensional space of T12 is obtained from the knowledge base 22 T12 of T12, for example by statistical inference on this knowledge base using the determined characteristics (position, orientation and shape parameter) as estimators.
  • This step which is described for vertebra T12 could be automatically reproduced for all the objects to be reconstructed, including object L1, in particular its lower end plate, in order to obtain the subject-specific reconstruction of the structure, a projection of which in the detection data plane is represented on FIG. 8 c.
  • this method will provide an accurate subject-specific reconstruction for portions of objects neighbouring the subject-specific reconstruction of the originally modified object. This method could be performed a given number of times, on different portions of objects, until a satisfying subject-specific reconstruction for the structure as a whole is obtained.
  • each modified preliminary solution could provide a set of parameters, comprising one or more parameters which are used to estimate a set of parameters, comprising one or more parameters, for the subject-specific reconstruction of the second object.
  • the subject-specific reconstruction obtained using the above-described method could be, in a last step, further refined by applying a non-linear rigid transformation such as kriging, such as for example described in Trochu, “A contouring program based on dual kriging interpolation”, Engineering with Computers, 9(3), 160-177, 1993, or any other suitable method.
  • kriging such as for example described in Trochu, “A contouring program based on dual kriging interpolation”, Engineering with Computers, 9(3), 160-177, 1993, or any other suitable method.
  • the method which was described above was by way of example only.
  • the structure to be studied is not necessarily the spine, but could be any part of the body skeleton of a patient, either in a lying or standing position.
  • the method was described using radiographs as a detection data.
  • any other kind of suitable detection data either bi-dimensional or three-dimensional could be used within the scope of the invention.
  • the data used is even not necessarily detection data, but could for example be any subject-specific data.
  • the initial solution of the subject-specific reconstruction was a three-dimensional initial solution, but the invention is not intended to be limited to the specific embodiment, and a three-dimensional subject-specific reconstruction could for example be obtained from a bi-dimensional initial solution.
  • the acquisition scheme is not limited to the one presented in relation to FIG. 1 , wherein a lateral and an antero-posterior images of the structure are obtained simultaneously.
  • the platen 29 will be movable relative to the radiological source in order to take a plurality of images of the patient standing on the platen along different orientations.

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Geometry (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Computer Graphics (AREA)
  • Software Systems (AREA)
  • Apparatus For Radiation Diagnosis (AREA)
  • Image Processing (AREA)
US12/601,975 2007-05-25 2007-05-25 Method for reconstruction of a three-dimensional model of a body structure Active 2030-05-25 US8660329B2 (en)

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/IB2007/001383 WO2008146069A1 (fr) 2007-05-25 2007-05-25 Procédé de reconstruction de modèle tridimensionnel de structure corporelle

Publications (2)

Publication Number Publication Date
US20100174673A1 US20100174673A1 (en) 2010-07-08
US8660329B2 true US8660329B2 (en) 2014-02-25

Family

ID=38442029

Family Applications (1)

Application Number Title Priority Date Filing Date
US12/601,975 Active 2030-05-25 US8660329B2 (en) 2007-05-25 2007-05-25 Method for reconstruction of a three-dimensional model of a body structure

Country Status (3)

Country Link
US (1) US8660329B2 (fr)
EP (1) EP2149123B1 (fr)
WO (1) WO2008146069A1 (fr)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20110211045A1 (en) * 2008-11-07 2011-09-01 Telecom Italia S.P.A. Method and system for producing multi-view 3d visual contents
US20160354161A1 (en) * 2015-06-05 2016-12-08 Ortho Kinematics, Inc. Methods for data processing for intra-operative navigation systems
US11707203B2 (en) 2016-10-11 2023-07-25 Wenzel Spine, Inc. Systems for generating image-based measurements during diagnosis

Families Citing this family (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20090226055A1 (en) * 2004-12-10 2009-09-10 Harry Dankowicz Systems and methods for multi-dimensional characterization and classification of spinal shape
EP2056255B1 (fr) 2007-11-02 2019-07-03 Ecole Nationale Supérieure d'Arts et Métiers (ENSAM) Procédé de reconstruction d'un modèle tridimensionnel d'une structure ostéo-articulaire
ES2382774B1 (es) 2010-02-12 2013-04-26 Universitat Pompeu Fabra Metodo para obtener una reconstruccion tridimensional a partir de una o mas vistas proyectivas, y uso de la misma
EP3150113A1 (fr) * 2010-12-13 2017-04-05 Ortho Kinematics, Inc. Procédés, systèmes et dispositifs de rapport de données cliniques et de navigation chirurgicale
GB201107385D0 (en) * 2011-05-04 2011-06-15 Materialise Nv Medical imaging calibration device
US20140282356A1 (en) * 2013-03-15 2014-09-18 SimuQuest, Inc. System Integration Techniques
US10524723B2 (en) * 2014-07-23 2020-01-07 Alphatec Spine, Inc. Method for measuring the displacements of a vertebral column
FR3035785B1 (fr) 2015-05-07 2017-06-02 Arts Procede d'estimation de la repartition de la densite minerale osseuse dans au moins une partie de squelette d'un individu
US11141221B2 (en) * 2015-11-19 2021-10-12 Eos Imaging Method of preoperative planning to correct spine misalignment of a patient
CN109584351B (zh) * 2017-09-29 2023-06-13 大连恒锐科技股份有限公司 一种基于生物特征的人身立体重建方法和系统

Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20010026637A1 (en) * 2000-01-14 2001-10-04 Pierre Lelong Processing method and system for 3-D geometric modeling of the spine
US6621491B1 (en) * 2000-04-27 2003-09-16 Align Technology, Inc. Systems and methods for integrating 3D diagnostic data
FR2856170A1 (fr) 2003-06-10 2004-12-17 Biospace Instr Procede d'imagerie radiographique pour la reconstruction tridimensionnelle, dispositif et programme d'ordinateur pour mettre en oeuvre ce procede
US20050004451A1 (en) 2002-04-26 2005-01-06 Stefan Vilsmeier Planning and navigation assistance using two-dimensionally adapted generic and detected patient data
US20060110037A1 (en) * 2002-07-19 2006-05-25 Michael Kaus Simultaneous segmentation of multiple or composed objects by mesh adaptation
US20060204069A1 (en) * 2002-12-20 2006-09-14 Biospace Instruments Method and device for radiographic imaging
US7194295B2 (en) 2002-03-27 2007-03-20 Brainlab Ag Medical navigation and/or pre-operative treatment planning with the assistance of generic patient data
US20070093998A1 (en) 2002-12-30 2007-04-26 Fouad El-Baroudi Method for biomehcanically simulating of a set of osseous joints
US7227981B1 (en) 1999-09-17 2007-06-05 Universite Joseph Fourier Three-dimensional statistic reconstruction of surfaces

Patent Citations (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7227981B1 (en) 1999-09-17 2007-06-05 Universite Joseph Fourier Three-dimensional statistic reconstruction of surfaces
US20010026637A1 (en) * 2000-01-14 2001-10-04 Pierre Lelong Processing method and system for 3-D geometric modeling of the spine
US6621491B1 (en) * 2000-04-27 2003-09-16 Align Technology, Inc. Systems and methods for integrating 3D diagnostic data
US7194295B2 (en) 2002-03-27 2007-03-20 Brainlab Ag Medical navigation and/or pre-operative treatment planning with the assistance of generic patient data
US20050004451A1 (en) 2002-04-26 2005-01-06 Stefan Vilsmeier Planning and navigation assistance using two-dimensionally adapted generic and detected patient data
US20060110037A1 (en) * 2002-07-19 2006-05-25 Michael Kaus Simultaneous segmentation of multiple or composed objects by mesh adaptation
US20060204069A1 (en) * 2002-12-20 2006-09-14 Biospace Instruments Method and device for radiographic imaging
US20070093998A1 (en) 2002-12-30 2007-04-26 Fouad El-Baroudi Method for biomehcanically simulating of a set of osseous joints
FR2856170A1 (fr) 2003-06-10 2004-12-17 Biospace Instr Procede d'imagerie radiographique pour la reconstruction tridimensionnelle, dispositif et programme d'ordinateur pour mettre en oeuvre ce procede
US20050008219A1 (en) * 2003-06-10 2005-01-13 Vincent Pomero Method of radiographic imaging for three-dimensional reconstruction, and a computer program and apparatus for implementing the method
US7639866B2 (en) 2003-06-10 2009-12-29 Biospace Med Method of radiographic imaging for three-dimensional reconstruction, and a computer program and apparatus for implementing the method

Non-Patent Citations (9)

* Cited by examiner, † Cited by third party
Title
Benameur et al., "3D/2D Registration and segmentation of scoliotic vertebrae using statistical models, Computerized Medical Imaging and Graphics", Pergamon Press, New York, NY, US, vol. 27 No. 5., Sep. 2003, pp. 321-337.
Benameur et al., "A hierarchical statistical modeling approach for the unsupervised 3-D biplanar reconstruction of the scoliotic spine", IEEE Transactions on Biomedical Engineering IEEE USA, vol. 52 No. 12, Dec. 2005.
Deschenes et al. ("3D Reconstruction of the Human Spine from bi-planar radiographs: using multi-scale wavelets analysis and splines interpolators for semi-automation," Proceedings of SPIE, vol. 5032 (2003), pp. 754-761). *
English Abstract of FR 2856170, (2004).
English Abstract of FR 2856170.
Kadoury et al. ("Personalized X-Ray 3-D Reconstruction of the Scoliotic Spine From Hybrid Statistical and Image-Based Models," IEEE Transaction on Medical Imaging, vol. 38, No. 9, Sep. 2009, pp. 1422-1435). *
Said Benameur, "Reconstruction 3D Biplanaire non Supervisee de la colonne vertebrale et de la cage thoracique scoliotiques par modeles statistiques", Ecole de Technologies Superieure Universite du Quebec, Montreal, Quebec, www.iro.umontreal.ca/{benameus/Publications/thesus-benameur.pdf, (2004).
Said Benameur, "Reconstruction 3D Biplanaire non Supervisee de la colonne vertebrale et de la cage thoracique scoliotiques par modeles statistiques", Ecole de Technologies Superieure Universite du Quebec, Montreal, Quebec, www.iro.umontreal.ca/{benameus/Publications/thesus-benameur.pdf,.
Search Report from PCT/IB/2007/001383 dated Sep. 21, 2007.

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20110211045A1 (en) * 2008-11-07 2011-09-01 Telecom Italia S.P.A. Method and system for producing multi-view 3d visual contents
US9225965B2 (en) * 2008-11-07 2015-12-29 Telecom Italia S.P.A. Method and system for producing multi-view 3D visual contents
US20160354161A1 (en) * 2015-06-05 2016-12-08 Ortho Kinematics, Inc. Methods for data processing for intra-operative navigation systems
US10959786B2 (en) * 2015-06-05 2021-03-30 Wenzel Spine, Inc. Methods for data processing for intra-operative navigation systems
US20210220057A1 (en) * 2015-06-05 2021-07-22 Wenzel Spine, Inc. Surgical navigation processors and systems
US11707203B2 (en) 2016-10-11 2023-07-25 Wenzel Spine, Inc. Systems for generating image-based measurements during diagnosis

Also Published As

Publication number Publication date
EP2149123A1 (fr) 2010-02-03
WO2008146069A1 (fr) 2008-12-04
US20100174673A1 (en) 2010-07-08
EP2149123B1 (fr) 2013-06-05

Similar Documents

Publication Publication Date Title
US8660329B2 (en) Method for reconstruction of a three-dimensional model of a body structure
US7639866B2 (en) Method of radiographic imaging for three-dimensional reconstruction, and a computer program and apparatus for implementing the method
US9142020B2 (en) Osteo-articular structure
US7095881B2 (en) Processing method and system for 3-D geometric modeling of the spine
EP3486873B1 (fr) Détection automatique d'implant à partir des artefacts d'image
EP2245569B1 (fr) Procédé d'imagerie médicale et système destiné à fournir un modèle d'élément fini
US20170273614A1 (en) Systems and methods for measuring and assessing spine instability
JP7092190B2 (ja) 画像解析方法、セグメンテーション方法、骨密度測定方法、学習モデル作成方法および画像作成装置
KR20120123370A (ko) 환자 척주의 특징을 측정하는 장치 및 방법
EP3355268A1 (fr) Procédé, dispositif de calcul et système de mesure d'une radiographie d'une zone d'examen médicale
Cheriet et al. Towards the self-calibration of a multiview radiographic imaging system for the 3D reconstruction of the human spine and rib cage
Giannoglou et al. Review of advances in Cobb angle calculation and image-based modelling techniques for spinal deformities
van Eeuwijk et al. A novel method for digital X-ray imaging of the complete spine
KR20160126195A (ko) 컴퓨터를 이용한 x-선 영상처리 골 모델링 작성방법, 시스템 및 컴퓨터 프로그램
US20230071033A1 (en) Method for obtaining a ct-like representation and virtual x-ray images in arbitrary views from a two-dimensional x-ray image
EP3756546A1 (fr) Indice de trabécules osseuses pour radiographie en champ sombre aux rayons x
Debrunner et al. Tomographic reconstruction from an uncontrolled sensor trajectory
CN117615731A (zh) 通过成像和跟踪验证脊柱弯曲矫正的方法和系统
Pomero et al. Fast semiautomatic stereoradiographic reconstruction of scoliotic spines using multi-scale image processing and statistical geometric models
Penney et al. Deforming a preoperative volume to better represent the intraoperative scene
Sadowsky et al. Enhancement of mobile c-arm cone-beam reconstruction using prior anatomical models
Bifulco et al. Research Article 2D-3D Registration of CT Vertebra Volume to Fluoroscopy Projection: A Calibration Model Assessment
de Moura Unsupervised Geometrical Modelling of the Scoliotic Spine from Radiographs
Pierre LELONG et al. 172 Research into Spinal Deformities 3 A. Tanguy and B. Peuchot (Eds.) IOS Press, 2002

Legal Events

Date Code Title Description
AS Assignment

Owner name: ECOLE DE TECHNOLOGIE SUPERIEURE, CANADA

Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:DE GUISE, JACQUES;GODBOUT, BENOIT;PARENT, STEFAN;REEL/FRAME:025990/0910

Effective date: 20101130

Owner name: CENTRE NATIONAL DE LA RECHERCHE SCIENTIFIQUE (CNRS

Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:HUMBERT, LUDOVIC;REEL/FRAME:025991/0348

Effective date: 20101117

Owner name: ECOLE NATIONALE SUPERIEURE D'ARTS ET METIERS (ENSA

Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:HUMBERT, LUDOVIC;REEL/FRAME:025991/0348

Effective date: 20101117

Owner name: ECOLE NATIONALE SUPERIEURE D'ARTS ET METIERS (ENSA

Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:SKALLI, WAFA;MITTON, DAVID;DUBOUSSET, JEAN;SIGNING DATES FROM 20110202 TO 20110214;REEL/FRAME:025991/0396

STCF Information on status: patent grant

Free format text: PATENTED CASE

MAFP Maintenance fee payment

Free format text: PAYMENT OF MAINTENANCE FEE, 4TH YEAR, LARGE ENTITY (ORIGINAL EVENT CODE: M1551)

Year of fee payment: 4

MAFP Maintenance fee payment

Free format text: PAYMENT OF MAINTENANCE FEE, 8TH YEAR, LARGE ENTITY (ORIGINAL EVENT CODE: M1552); ENTITY STATUS OF PATENT OWNER: LARGE ENTITY

Year of fee payment: 8